Phonak's "Audiogram Direct"

I understand that many Phonak models when programmed using their iPFG software give the fitter an option of using a function called “Audiogram Direct.”

As I understand it, this enables the the instrument to generate its own (internal) calibrated test “tones” at the frequencies and levels the fitter chooses. The user then responds with his/her response as to whether it sounded absent, quiet, medium level or loud. The result (threshold) is stored in the fitter’s program (iPFG) and the fitter can use it as he sees fit.

This sounds useful. Does anyone use this feature? If so please give us the details, your test method, and your opinion of this feature. It could be used as a starting point. As a broad-brush diagnostic tool it might detect if your instrument is defective, the original audiogram was off or what frequencies to tweak in fine tuning.

Our paediatric department uses it with all children’s fittings where a child is old enough to respond, which would imply they think it’s a useful verification of a fitting. Anyway, thanks for explaining what it does, I have seen it on iPFG but cos I don’t have any hearing aids connected I had no idea what it was meant to do!.

As a self programmer I used it to create an audiogram. This was then used to set up the hearing aids. I found the audiogram done via audigramdirect better suited to my hearing loss than the one done by a professional in a sound room.


I’m a self-programmer too and take an iterative approach to refining performance.

I wear Savias (pre-AudiogramDirect). I programmed them here using audiograms done conventionally along with ones I made from the tests at using Rube Goldberg artificial ears, sound meters, calibrated headsets etc… (So far the ones done by our local audiologists vary wildly.*) Sad to say at this point I don’t trust the ones done conventionally any more than my own (in other words both are questionable).

The Savia BTE’s still work pretty well but now I want to augment the Savias with more recent Phonaks and then use AudiogramDirect to refine/replace the audiogram data I use, using the same (pretty much occluding) molds.

Thanks for the responses.

  • I know there are thousands of true professionals who do truly excellent work but unfortunately I haven’t located them yet in this area.

Hey, those were really interesting online tests, thank you. I always thought my very low frequency hearing was poor but this confirms I have no responses at all below the standard starting point for an audiogram, not a thing. Bass boost and such like on hi-fi systems just makes me feel sick because it shakes the floor and I can’t hear it!

But what was really interesting was that on some of the tones I did hear something but not at all what one would expect to hear. For example, on the 1kHz tone I heard as far as 60dB on the R side but when I clicked for the 55 or under I still heard a noise, I would say it was pretty much exactly an octave above the 1kHz tone, so there was a noise, but it was not a 1kHz tone! I get this a lot in hearing tests, I can hear, feel or otherwise sense something, but it’s definitely not the sound I am being asked to hear. Weird!!

Here are two audiograms. The first was generated after using the online hearing test which suggest very little hearing loss in the mid range.

The second is done using audiogramdirect. This configuration I have been using for several months.

I have not had a chance to use the online test in real life. Initial impression is that the gain used is no were’s near sufficent for my hearing loss.

I don’t understand why the audiograms have not been automated. Operator error by the audiologist seems to be an issue at times…

It appears that this feature is not in the Versata. :mad:

What do you mean by not automated?


You said: "I have not had a chance to use the online test in real life. Initial impression is that the gain used is no were’s near sufficient for my hearing loss. "

Remember the AudiogramDirect result is also affected by the actual attenuation of the tube/mold/ear-canal.

The fact that AudiogramDirect produced an audiogram with much higher loss at the higher frequencies for you could mean that the original audiogram was off OR it could mean that your tube/mold/ear-canal section actually has higher attenuation at the higher frequencies than iPFG thinks it should have had.

You might want to trouble-shoot/replace the tube/mold section or try a different tube/mold approach.

I wear a pair of Audeo smart V’s with standard power receivers.

It only took me a few minutes outdoors to know that the online hearing test did not meet my needs for gain.

Sorry I mixed up the posts/names. I guess you were referring to the test? If so I think your assessment is probably right.

As you know this test is not calibrated at all until/unless you nail it with suitable earphones calibrated with a sound meter w “ear” attachment per their instructions. Even then accuracy is probably not going to be great.

AudiogramDirect is an attractive option in my case.

I have to say the Audiogramdirect is very nice. I had seen it but never used it. The audiogram from the audioolgist was off by 10 to 20 db. My Exelia’s programming is now recalculated per the Audiogramdirect with much better results.

Thanks for this post.:slight_smile:

From what I can tell, all Versata models have AudiogramDirect.

this is not meant to replace conv. audiogram, remember no Bone conduction
no masking. It could be a good tool to do a quick finetunning

Thanks for pointing out the limitations. The audiogram has important purposes aside from just setting up the hearing instrument.

As you know it is very common that a user will buy sophisticated Phonak instruments from a dispenser very unfamiliar with them and the dispenser is reluctant to move much beyond the factory (default?) settings and the simplest most cautious adjustments. If the tube/mold/canal characteristics are not what iPFG guesses it to be (based on perhaps inexpert inputs), and the dispenser is reluctant to get in deeper, then the user either returns the unit or suffers with its limitations for years. You hear this story over and over - it is not a very uncommon situation.

As you pointed out AudiogramDirect might be a good tool to uncover such mistakes - bad or sloppy audiograms (like the ones I received) incorrect entries about venting, invalid approximations about the user’s canal etc. Audiogramdirect it seems to me is a chance to throw the cards in the air, ignore what’s been entered and see if the result turns out to be significantly different.

At minimum it’s a powerful troubleshooting procedure that could at least tell the user, the dispenser (or a remote Phonak rep) that mistakes/problems were present in the first pass at programming it and that it is still possible for this user to get an acceptable result.

You stress that AudiogramDirect is not a complete solution. Ideal or not, in some difficult situations where accuracy, troubleshooting expertise and/or motivation are lacking or where getting a thorough, precise audiogram is not in the cards, AudiogramDirect applied by a dispenser or a self-programmer may be the only viable route available to the user to make effective use of his investment or to prevent an unwise investment.

It’s not there to suggest sloppiness on the part of the dispenser, it’s to take account of the different canal impedances caused by the unique characteristics of your ear canal. It’s a way of taking into account the ideas behind a REM without actully performing it - at least for threshold. It’s been on other software for ages.

Any venting on the aid will pretty much invalidate the audiogram below 1Khz as there will be too much energy loss to obtain an accurate threshold measurement.

IMHO, it ought never to be used in the absence of a full audiogram - least of all as you won’t have any basis for the initial aid selection.

This has been a great thread. Thanks for all the info.

The lower frequency db readings are in fact showing much more hearing loss as umbongo says. With my audiogram the difference is about 25 db difference at 250 Hz and slowly tapers to the same db reading as the audiologist audiogram at 4000 Hz. I could not get any more readings after 100 db on the Audiogramdirect.

Using the Audiogramdirect has improved my hearing, particularly speech recognition and the TV.

Thanks again.

Perhaps not. But what should ideally happen often isn’t what is actually available to the user.

Once you have the instrument and when re-tweaking has run its course (due to whatever factors) then you are faced with turning to trouble-shooting procedures, living with unacceptable performance or throwing the instruments in a drawer and starting again. I choose the trouble shooting route. AudiogramDirect has a possibility of exposing mistakes or omissions.

Whether or not you would end up using the audiogram it generates, I’d like to hear that dispensers use AudiogramDirect effectively as a trouble-shooting tool or for double-checking. I wouldn’t hesitate to use it that way. There is no shame in double-checking your work.

It’s been around for years on Widex product as the Sensogram, Bernafon etc. I tend to use it heavily on home visit dispenses as I don’t carry a portable REM kit. If your dispenser has any sense, they will have run the test on or just after the first fit, as it allows the fitting to take into account the residual ear canal volume, flexibility of the ear-drum etc.

It’s not a silver-bullet and can be incredibly misleading for the low frequency results, BUT like you said, if the results of your initial fitting are poor, it will be a great help in giving you an idea of threshold and appropriate gain. However, as you still aren’t in control of the real-time performance of the aid, you have to use the results with caution - especially as you are still reliant on manufacturer prescribed target output based on the new Audiogram.

It seems like ‘hit and hope’ dispensing is still alive and well in some areas… :rolleyes: